On Thu, Feb 14, 2013 at 12:40 PM, Christophe Fergeau <cfergeau@xxxxxxxxxx> wrote: > On Wed, Feb 13, 2013 at 07:16:59PM +0200, Zeeshan Ali (Khattak) wrote: >> On Wed, Feb 13, 2013 at 6:54 PM, Christophe Fergeau <cfergeau@xxxxxxxxxx> wrote: >> > I'm still not fully convinced that we really need all this 'format' API in >> > the first place, pre/post install drivers and install scripts are all >> > listed in an <os> node, if an install script for a given OS does not >> > support a driver format, why is it listed in the first place in the <os> >> > node? >> >> I just wanted to support the case of providing information about >> drivers even if our autoinstallation script does not support them. >> Having said that I'm not sure such information would be of any use to >> the apps. No strong opinions so I'll let you decide if its needed or >> not. > > At worse (ie if in the future we add an install script that does not > support one driver we list), applications will try to use the driver, but > this will fail, right? I guess when we get to such a situation, we can look > into adding this API, but that for now having it is not required. Correct > me if I'm missing something :) Actually you and I both missed something important here: The libosinfo db doesn't just come from libosinfo. We allow downstream and apps to add to the DB. They can as well add new device drivers. Keeping that in mind, I would like you to reconsider the format related patches. -- Regards, Zeeshan Ali (Khattak) FSF member#5124 _______________________________________________ Libosinfo mailing list Libosinfo@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libosinfo
